Yesterday, in an expert interview, I was asked whether the success factors in dealing with AI applications differ from those that have played an important role in employees’ working lives to date. I found the question extremely interesting. I came to the conclusion that the success factors do not differ significantly, but that they have an even more pronounced effect.
Sundar Pichai, CEO of Google, is credited with saying that AI is probably the most important thing humanity has ever worked on. It is at least as fundamental as electricity and has the potential to improve our world. In fact, AI is already proving to be a tremendous source of value creation. Proven methods are no longer enough. We need to engage intensively with AI in order to integrate it meaningfully into our operational processes. So, what are the crucial success factors in using KI applications?
According to my insights, people with good leadership skills will also be more successful with AI applications than those who do not have good leadership skills. The reason for this is that AI delivers better results
- the more structured the approach is,
- the more carefully KI applications are chosen for a specific task,
- the more clearly their roles and functions are defined,
- the better the chosen applications are interconnected,
- the more focused and precise their instructions are,
- and the more clearly the context for a specific task to be performed is described..
Just as it is crucial in human management structures to recognize which individuals are capable of performing a task well, AI applications must also be carefully selected, as they too have certain capabilities and limitations. And their collaboration should be enabled to secure the workflow quality. In addition, evaluation and feedback on the results delivered are important in order to enable the chosen AI applications to revise and learn. The more coordinated the dialogue with AI applications is, the more effectively the AI will support and relieve its users.
These are all management issues that we are already familiar with from previous practice without AI. Employees also need feedback and appreciation for orientation and as a motivating source of meaning. Now we could also mention netiquette. It has actually been shown that AI applications deliver higher-quality results when users politely ask them for support and explain how important the results are for achieving their personal goals. Of course, an AI application is an emotionless machine, but it takes from the wording the information that it should make a special effort; netiquette thus also generates something like motivation in AI applications.
IT affinity is often cited as a prerequisite for the effective use of AI. I accept this objection to a limited extent, in that users of AI applications should have experience with PCs. Many intuitively operable AI applications are now available that do not require any special IT knowledge, especially programming skills.
The challenge remains to embrace new AI applications and understand their logic. This requires a dose of curiosity and the unconditional will to persevere and “stick with it.” In this respect, too, I see parallels with the working world as we know it today. These are expectations that were already among the factors for success in the world without AI.
Finally, it is important to consider the tasks for which AI should be used. These are strategic, creative, and leadership issues that we are also familiar with from the world we knew before. Ask yourself: Can an AI application do what I do faster, better, and with fewer errors? Which AI-based solution contributes to the company’s benefit? Which business cases are relevant for AI support? Which processes should be automated using multi-agent tools? How can AI not only increase efficiency, but also scale sales and revenue through AI-based smart workflows?
The success factors in dealing with AI applications do not differ significantly from the success factors without AI, but with AI, the extent to which the success factors are pronounced is likely to have a greater impact on the outcome.